README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: duration
      dtype: float64
    - name: sentence
      dtype: string
    - name: uid
      dtype: string
    - name: group_id
      dtype: string
  splits:
    - name: train
      num_bytes: 5562218134.16
      num_examples: 20519
    - name: valid
      num_bytes: 628281897.6
      num_examples: 2565
    - name: test
      num_bytes: 636312103.8
      num_examples: 2565
  download_size: 6552336388
  dataset_size: 6826812135.56
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: valid
        path: data/valid-*
      - split: test
        path: data/test-*